7 root effects on soil organic matter decomposition - …kuzyakov/cheng_asa_2005_roots-som.pdf ·...

Root Effects on Soil Organic Matter Decomposition

WEUXUB CHENG

University of CaliforniaSanta Cruz, California

YAKOV KUZYAKOV

University of HoheinhemStuttgart, Germany

Roots of higher plants, a key functional component of belowground systems andone of the main soil-forming agents (Jenny, 1941), interact with virtually all soilcomponents. Processes that are largely controlled or directly influenced by rootsand often occur in the vicinity of the root surface are often referred to as rhizo-sphere processes. These processes may include root production through growthand death (root turnover), rhizodeposition, root respiration, and rhizosphere mi-crobial respiration that is a result of microbial utilization of rhizodeposits. Rhi-zosphere processes play an important role in the global C cycle (Fig. 7–1). Ter-restrial ecosystems are intimately connected to atmospheric carbon dioxide (CO2)levels through photosynthetic fixation of CO2, sequestration of CO2 into biomassand soils, and the subsequent release of C through respiration and decompositionof organic matter. Considering all the pools and fluxes of C within ecosystems, Ccycling belowground is increasingly being recognized as one of the most signifi-cant components (Jackson et al., 1997; Zak and Pregitzer, 1998). Globally, theinput of C to the soil has been estimated to be as great as 60 ( 10−5 g yr−1 (Post etal., 1990). Thus, small changes in the equilibrium between inputs and decomposi-tion could have a significant impact on atmospheric CO2 concentrations, whichmay either exacerbate or reduce the consequence of burning of fossil fuels(Schimel, 1995).

Carbon dioxide efflux from the soil is a combination of two distinctprocesses: (i) rhizosphere respiration, including root respiration and microbialrespiration from the metabolism of rhizodeposits and (ii) microbial decompositionof soil organic matter (SOM). The substrates for rhizosphere respiration comefrom C recently fixed through photosynthesis, whereas SOM decomposition isprimarily a function of soil heterotrophic activities using soil C. The two processesact simultaneously and are also linked through rhizosphere interactions, which

Kitty’s TS • American Society of Agronomy • Wright & Zobel . . . • 183225

7

Copyright © 2005. American Society of Agronomy, Crop Science Society of America, Soil ScienceSociety of America, 677 S. Segoe Rd., Madison, WI 53711, USA. Roots and Soil Management: Inter-actions between Roots and the Soil, Agronomy Monograph no. 48.

119

07Cheng 5/20/05 13:34 19


120 CHENG & KUZYAKOV

Fig. 7–1. A conceptual model depicting potential mechanisms controlling rhizosphere C dynamics.The unit of the numbers is 1015 g (annually for fluxes) (Based on Post et al., 1990).

may exert a stimulative (priming effect) or a suppressive influence on SOM de-composition (Cheng, 1999; Van Veen et al., 1991). The focus of this chapter is onthis linkage and the associated rhizosphere interactions.

Large amounts of C and mineral nutrients contained in plants recycle backto the soil through rhizosphere processes (Pregitzer et al., 1995). A significantamount of work has been done quantifying root production (Vogt et al., 1986;Cheng et al., 1990; Hendrick and Pregitzer, 1993; Eissenstat and Yanai, 1997) andtotal rhizodeposition (Lynch and Whipps, 1990; Kuzyakov and Domanski, 2000).However, only limited effort has been made to link rhizosphere processes with soilprocesses such as organic matter decomposition and nutrient transformation. A re-cent report from the EUROFLUX project clearly illustrates the potential impor-tance of rhizosphere processes in determining the net C gain or loss in 18 forestecosystems in Europe (Janssens et al., 2001). It is challenging to link rhizosphereprocesses with soil processes such as organic matter decomposition. Althoughstudies (e.g., Helal and Sauerbeck, 1984, 1986; Liljeroth et al., 1994) have indi-cated that input of labile substrates in the rhizosphere may significantly enhanceSOM decomposition as a result of the priming effect (Dalenberg and Jager, 1989),rates of SOM decomposition are commonly assessed by laboratory incubations ofsoil samples with an often implicit assumption that rhizosphere processes have lit-tle impact on the results. However, the assumption has rarely been rigorouslytested. Many questions remain to be answered. What is the potential effect of therhizosphere on SOM decomposition? Is it significant enough to warrant serious in-vestigation? How does the rhizosphere effect on SOM decomposition changethrough time? What biotic and abiotic factors (e.g., N, CO2, light, soil types, andplant species) control or influence the level of root effects on SOM decomposi-tion? Which kinds of rhizosphere mechanisms and interactions change the SOM

07Cheng 5/20/05 13:34 Page 120


ROOT EFFECTS ON SOIL ORGANIC MATTER DECOMPOSITION 121

Fig. 7–2. The separation of different sources of carbon dioxide (CO2) evolved from the belowgroundsystem.

decomposition rate? By reviewing published results on this issue, we aim to ad-dress these questions and describe the current stage of understanding on the mech-anisms that regulate these root effects.

METHODOLOGY

Studies of root effects on decomposition processes have been restricted bythe limitation of existing methods. Carbon dioxide released by a respiring systemof living roots and soil may have four origins (Fig. 7–2): (i) root respiration, (ii)microbial respiration using substrates from live roots (rhizo-microbial respira-tion), (iii) rhizosphere-stimulated (or suppressed) microbial respiration usingSOM as the substrates, and (iv) microbial respiration using SOM without the in-fluence of live roots (basal respiration), which is often measured in soil incuba-tions. The third source is also called the “primed decomposition” or the “primingeffect.” Total rhizosphere respiration is defined as the sum of root respiration andrhizo-microbial respiration. Carbon used in total rhizosphere respiration is all de-rived from photosynthesis or storage of living plants. Studying root effects on de-composition requires separation of decomposing sources of focus from sources oflive roots. This has been mostly accomplished by using isotope techniques. Meth-ods used in published studies can be grouped into five main categories: (i) labelled

07Cheng 5/20/05 13:34 Page 121



litter, (ii) pulse-labelling of plants, (iii) continuous labelling of plants, (iv) natural13C tracers, and (v) N budgeting. Without C isotopic labelling and tracing, simpleC budgeting approach is inadequate for studying root/rhizosphere effects on SOMdecomposition because of the unavoidable mixing of C sources from roots andSOM. However, N budgeting has been used to evaluate the effect of roots on Nmineralization with considerable success (e. g., Wang and Bakken, 1997; Merbachet al., 1999). Because the quantity of N deposition from living roots is often neg-ligible compared to the amount of soil mineral N or N taken up by plants, total Nmineralization can be a reasonable measure of SOM decomposition if microbialN immobilization is also negligible.

With the labelled-litter method, plants are grown in soils mixed with iso-tope-labelled (C or N) litter, so that the release of the labelled source between theplanted treatment and an unplanted control can be compared. The difference in ei-ther CO2 production or N release allows the determination of root effects on therate of litter decomposition if other soil conditions have been kept the same. Thismethod requires production of uniformly labelled litter before the start of the ex-periment. Some studies indicate that multiple pulse labelling is an acceptablealternative to continuous labelling to produce labelled litter for decompositionstudies (Sparling et al., 1982; Sallih and Bottner, 1988). However, the degree ofuniformity of the labelled litter has been an issue of concern because unevenly la-belled litter may produce seriously biased results. Litter labelled with 14C has beenthe most common form in published studies. This method is relatively easy to use.The major drawback of this method is that it only measures the change in the la-belled components in litter, and does not necessarily represent the decompositionof SOM.

The pulse labelling method involves pulse-labelling plants with C isotopesso that C from the live roots can be separately monitored from the soil C (Kuzyakovet al., 2001; Kuzyakov and Cheng, 2001). Difference in the rates of soil C loss be-tween the planted treatment and the unplanted control is a measure of root effecton the rate of SOM decomposition. This method has not been used often becauseit involves complicated calculations and some assumptions (Kuzyakov et al.,2001). One of the assumptions is that the specific activity of the root-derived CO2is the same as that of the roots sampled at the end of the experiment, which has notyet been critically tested. Another assumption is that plant shoots and roots growlinearly during the period of the experiment. A comparison of the pulse-labellingmethod of Kuzyakov and Cheng (2001) with the natural 13C-tracer method ofCheng (1996) shows that the two methods may give similar results in terms of totalrhizosphere respiration measurements. However, because pulse labelling does notuniformly label all plant C, total plant-derived C cannot be separated from soil-derived C in a pulse-labelling experiment.

The continuous labelling method has been used in several studies (e.g., Lil-jeroth et al., 1994; Helal and Sauerbeck, 1983, 1984, 1986). This method requiresthat plants are grown in an atmosphere with labelled C dioxide (13CO2 or 14CO2) ofa relatively constant specific activity or enrichment throughout an entire experi-ment so that all plants are uniformly labelled, because photosynthetically fixed CO2is the sole source of plant C. This continuous labelling method allows separationof plant-derived C from the soil C. The difference in the rates of soil C loss between

07Cheng 5/20/05 13:34 Page 122



the planted treatment and the unplanted control is a measure of root effect on therate of SOM decomposition. This method focuses on the mineralization of total soilorganic C pool instead of on plant litter C only as in the case of the labelled littermethod. Both 14C pulse labelling and continuous labelling methods have limita-tions. Continuous 14C labelling requires special facilities that are available at onlya few places in the world. Also, the continuous 14C labelling method often requirestransplanting of seedlings which may have considerable unlabelled C reserves,therefore requiring some time for all plant parts to become uniformly labelled. Be-cause of the safety issue due to the use of radioactive materials, most 14C-labellingexperiments have been of short duration. The safety issue can be avoided if 14CO2is replaced with 13CO2 in the continuous labelling experiment. However, 13C massspectrometry analysis is much more expensive and time consuming than 14C scin-tillation counting.

Natural 13C abundance has also been used to trace root effects on SOM de-composition (Cheng, 1996; Qian et al., 1997; Rochette and Flanagan, 1997). Theprinciple of the natural 13C tracer method is based on two factors: (a) the differ-ence in the 13C:12C ratio (often reported in δ13C value) between plants with the C3photosynthetic pathway whose mean δ13C is �27‰ and plants with the C4 path-way whose mean δ13C is �12‰ (Smith and Epstein, 1971), and (b) the subse-quent difference between SOM derived from the two types of plants. SOM derivedfrom C4 plant-dominated vegetation (C4-derived soil) such as tallgrass prairies andtropical grasslands may have δ13C values ranging from �12 to �20‰, whereasδ13C values of SOM derived from cold and temperate forest (C3-derived soil) mayrange from �23 to �27‰. By using C4-derived soil in a C3 plant system or viceversa, the C entering the soil via live roots will have a different δ13C value than theδ13C value of SOM. Thus the total amount of C dioxide evolved from the root-soilsystem can be separated into the plant-derived source and the soil-derived source.

This natural 13C method eliminates some of the major limitations of earlierlabelling methods and offers a new opportunity for systematic studies of themechanisms regulating the quantity and quality of rhizodeposition as well as theinteractions between plant roots and SOM decomposition. As mentioned above,this method requires the switch of C3-plants from their original C3-soils to C4-derived soils or the switch of C4-plants from their original soils to C3-derived soils.Because of this switch, the composition of soil microbial communities may be dif-ferent from the original one. Therefore, one untested assumption associated withthis method is that the switches of plant species from their original soils do notsubstantially modify rhizosphere processes. Compared to the continuous 14C-labelling approach, a major drawback of this method is the much higher cost as-sociated with sample analysis for 13C (Cheng, 1996). Another limitation of thismethod is its relatively low sensitivity and low accuracy.

SYNTHESIS OF CURRENT RESULTS

The labelled-litter method has been one of the most common methods usedto study root effects on litter decomposition. This method was employed initiallyin several field studies (e.g., Jenkinson, 1977; Shields and Paul, 1973; Fuhr and

07Cheng 5/20/05 13:34 Page 123



Sauerbeck, 1968). In these field studies, the rate of loss of 14C-labelled C was mea-sured in treatments of planted and unplanted fallow control. One common resultof these studies was that planted treatments significantly reduced labelled litter de-composition as compared to the fallow control and that the soil moisture level inthe planted treatment was also significantly reduced. Therefore, the reduction inthe rates of litter decomposition was most likely caused by the drier soil conditionsdue to plant water uptake and transpiration. In other words, the effect of plantingand root growth on litter decomposition was confounded by their effects on soilwater conditions. In order to resolve this confounding outcome, soil water levelshave to be maintained to a similar level between the planted treatment and the un-planted control by frequent watering or irrigation. In this synthesis, we exclude re-sults generated from experiments that did not have adequate soil water control.

For comparisons between different studies, we calculated priming effectvalues as percents of the unplanted control (% priming) due to the presence of therhizosphere based on the following formula:

% priming = × 100

where RP is the SOM decomposition rate of the planted treatment and RNP is theSOM decomposition rate of the unplanted treatment. The presence of the rhizo-sphere stimulates SOM decomposition when % priming is a positive value andsuppresses decomposition when it is a negative value.

Labelled Litter Studies

The effect of roots on SOM decomposition, based on labelled litter methodsunder controlled water conditions, are summarized in Table 7–1. The % primingvalues from these studies ranged from 270% (Sparling et al., 1982) to 23% (Helaland Sauerbeck, 1987). These results demonstrated that the presence of roots mayeither enhance or suppress the decomposition of labelled litter, depending on thecoupling of plant species with soil types, experimental conditions, and the dura-tion of the experimental period (Sallih and Bottner, 1988). All these results wereobtained from experiments under well-watered and well-aerated conditions andwith limited plant types (all from the Monocotyledonae). The experimental dura-tion for the studies was <2 mo except for one study that lasted 690 d. Some stud-ies (Reid and Goss, 1982; Sparling et al., 1982) attempted to estimate root effectson SOM decomposition by incubating the labelled litter (shoots or roots) in thesoil before planting. However, the labelled materials from the litter would not beincorporated into the SOM evenly. Therefore, the results should be interpreted asroot effects on litter decomposition instead of SOM decomposition as a whole. Inthe report by Cheng and Coleman (1990), the influence of fertilization and micro-bial biomass was also assessed. They found that the amount of 14C-labelled mi-crobial biomass C was highly correlated with the amount of 14CO2 released dur-ing their experiment, and thereby speculated that microbial biomass or activitieswere the main determinant of the outcomes of root effect on decomposition. Helaland Sauerbeck (1987) also indicated that there was less 14C-labelled materials re-maining at the end of the experiment in the planted soil zones than soil zones awayfrom roots or the unplanted control. Sallih and Bottner (1988) noted that there was

RP � RNP

RNP

07Cheng 5/20/05 13:34 Page 124



Tabl

e 7–

1. M

agni

tude

of

the

root

eff

ect o

n th

e de

com

posi

tion

of la

bele

d lit

ter

adde

d to

the

soil.

Plan

t typ

eM

ater

ial o

f fo

cus

Soil

type

,C %

PGC

†%

Pri

min

g‡D

urat

ion(

d)R

efer

ence

Rye

14

C r

ye li

tter

Typi

c K

anad

ult,

1.1%

G

H11

49C

heng

and

Col

eman

(19

90)

Mai

ze14

C m

aize

litte

rL

uvis

ol/C

hern

ozem

GC

2325

Hel

al a

nd S

auer

beck

(19

87)

Mai

ze.

14C

bar

ley

root

sSa

ndy

loam

,2%

CG

H−5

322

Rei

d an

d G

oss

(198

2)L

oliu

m p

eren

ne14

C b

arle

y ro

ots

Sand

y lo

am,2

%C

GH

−38

42R

eid

and

Gos

s (1

982)

Mai

ze14

C b

arle

y ro

ots

Sand

y lo

am,2

%C

GH

−21

22R

eid

and

Gos

s (1

983)

Lol

ium

per

enne

14C

bar

ley

root

sSa

ndy

loam

,2%

CG

H−3

042

Rei

d an

d G

oss

(198

3)W

heat

14C

whe

at li

tter

Fers

ialli

tic c

alci

c,2.

7%C

GH

−10

to 2

069

0Sa

llih

and

Bot

tner

(19

88)

Bar

ley

14C

rye

litte

r4.

8% C

GH

−70

42Sp

arlin

g et

al..

(19

82)

†Pla

nt g

row

th c

ondi

tions

:GH

= g

reen

hous

e; G

C =

gro

wth

cha

mbe

r.‡E

ach

valu

e is

cal

cula

ted

as:(

plan

ted

2 un

plan

ted)

/unp

lant

ed ×

100.

07Cheng 5/20/05 13:34 Page 125



Table 7–2. Magnitude of the root effect on soil organic matter (SOM) decomposition as assessed bycontinuous 14C-labeling of plants.

Plant type Treatment Soil type, C % PGC† % Priming‡ Time (d) Reference

Wheat High N Loamy sand 2% C GC 33 54 Liljeroth et al. (1990)Maize High N Loamy sand 2% C GC 133 47 Liljeroth et al. (1994)Wheat High N Loamy sand 2% C GC 33 47 Liljeroth et al. (1994)Maize Low N Loamy sand 2% C GC 196 47 Liljeroth et al. (1994)Wheat Low N Loamy sand, 2% C GC 196 47 Liljeroth et al. (1994)Maize Chernozemic

sandy loam1.5% GC 236 25 Helal and Sauerbeck

(1984)Maize Chernozemic

sandy loam1.5% GC 332 30 Helal and Sauerbeck

(1986)

†Plant growth conditions: GH = greenhouse; GC = growth chamber.‡Each value is calculated as: (planted 2 unplanted)/unplanted × 100.

an inhibition effect of wheat (Triticum aestivum L.) roots during early period(0–200 d) of the experiment and a positive priming effect of roots during the laterpart (200–690 d) of the experiment. There were seven frequent harvests of shootsand roots during the whole experiment, which might have imposed a high degreeof disturbance to the experimental system. Results of the negative priming effectwere mostly reported by Reid and Goss (1982, 1983) and Sparling et al. (1982).These studies reported a significant amount of 14C-labelled C was taken up byplant roots, which might have contributed to the negative priming effect of theroots on decomposition of the labelled materials.

Continuous Labelling Studies

The magnitude of root effects on SOM decomposition was assessed in fourpublished studies using continuous labelling methods as found in a literature search(Table 7–2). In all these studies, soil water was controlled to a similar level betweenthe planted treatment and the unplanted control by frequent watering. The poten-tial effect of soil drying in planted treatments was at a minimum. The values of %priming were estimated using the data/figures reported in these studies. The %priming values ranged from 33% to as high as 332%. These results demonstratedthat the presence of roots enhanced the decomposition rate of original SOM in allfour studies. The experimental duration was short for all studies (<55 d). Nitrogenfertilization reduced the priming effect of roots on SOM decomposition rates from196% priming to 33% for wheat and from 196 to 133% for maize (Zea mays L.)(Liljeroth et al., 1994). Studies by Helal and Sauerbeck (1984, 1986) reported muchhigher % priming values (336 and 432%). These higher values were indirectly de-rived from soil C budgets (12C-C) instead of direct measurements of soil C loss as12CO2 as in the case of Liljeroth et al. (1990, 1994). The approach of total soil Cbudgeting has much lower sensitivity and accuracy than directly measuring soil Closs in the form of 12CO2 because the inherently high variability of total soil C con-tents often buries treatment differences in terms of soil C loss. However, the con-sistently lower amount of soil C remaining in the soil sections with roots or closer

07Cheng 5/20/05 13:34 Page 126



to the rooting zone than sections away from the rooting zone, also indicated a prim-ing effect by roots in their studies. Soil aggregate destruction by growing roots wasmentioned as the main cause of the priming effect by Helal and Sauerbeck (1987).These continuous labelling studies were able to assess the root effects on SOM de-composition, not just litter decomposition. However, the required use of specialcontinuous labelling facility might have prevented comprehensive investigation ofthe root priming effect both because of the short experimental duration permittedby such facility and because of the limited plant growth space.

Studies Using Natural 13C Tracers

The results of root effects on SOM decomposition were given in Table 7–3based on six published studies using natural 13C-tracer methods under controlledsoil water conditions. The % priming values were mostly positive, ranging from�37% to as high as 164%. In a very short (16 d) experiment using soils from acontinuous corn field, the presence of wheat roots was shown to reduce the SOMdecomposition rate by about 37% (Cheng, 1996). In another short (28 d) experi-ment (Cheng and Johnson, 1998), wheat plants grown under elevated CO2 in-creased rhizosphere priming effects on SOM decomposition when the soil re-ceived additional N, but decreased rhizosphere priming effects without N addition.Sunflower (Helianthus annuus L.) plants grown under elevated CO2 without N fer-tilization also produced less rhizosphere priming effects than ambient CO2 asshown by the results from a mesocosm-scale experiment lasting for 53 d (Chenget al., 2000). The highest degree of rhizosphere priming effects in this set was re-ported from the greenhouse study that lasted for 119 d (Cheng et al., 2003); therate of SOM decomposition under the influence of soybean [Glycine max (L.)Merr.] rhizosphere was 164% higher than the no-plant control and 96% higherwith the presence of wheat rhizosphere. Two kinds of plant-soil couplings (C3-plants grown in a “C4-soil” and C4-plants grown in a “C3-soil”) were compared inthe greenhouse study by Fu and Cheng (2002). The two C4 plant species grown inthe “C3-soil” (a coastal annual grassland soil) produced a lower rhizosphere effect(29 and 27%) on SOM decomposition than the two C3 plant species grown in the“C4-soil” (a tallgrass prairie soil). This difference could be due to the two soiltypes, to the different plant species with two different photosynthetic pathways, orboth. As has been pointed out by Kuzyakov et al. (2000), soils with higheramounts of labile SOM are more likely to produce priming effects than soils withless labile SOM. As indicated by its lower basal respiration rate, the coastal grass-land “C3-soil” probably contained less labile SOM than the tallgrass prairie soil(C4-soil). It is also possible that plant species with the C4-photosynthetic pathwayproduced less rhizosphere-priming than C3 plants (Epstein et al., 1998). However,similar rhizosphere priming effects were reported between maize (a C4 species)and wheat (a C3 species) when both received the low level of N fertilization (Lil-jeroth et al., 1994). In a controlled shading experiment (Kuzyakov and Cheng,2001), plant photosynthesis, as modified by different day-night cycles, was shownto exert an important control on rhizosphere priming effects. The SOM decompo-sition rate of the planted treatment under a regular day-night cycle (12 h light/12h dark) increased 100% above the no-plant control but decreased to a level of 50%

07Cheng 5/20/05 13:34 Page 127



Tabl

e 7–

3. M

agni

tude

of

the

root

eff

ect o

n so

il or

gani

c m

atte

r (S

OM

) de

com

posi

tion

by th

e na

tura

l 13C

trac

er m

etho

d.

Plan

t typ

eT

reat

men

tSo

il ty

pe,C

%PG

C†

% P

rim

ing‡

Tim

e (d

)R

efer

ence

Whe

atTy

pic

Frag

iuda

lfG

C−3

716

Che

ng (

1996

)W

heat

Am

bien

t CO

2,N

o-N

Mol

lisol

,cla

y lo

am,2

%G

C44

28C

heng

and

Joh

nson

(19

98)

Whe

atE

leva

ted

CO

2,N

o-N

Mol

lisol

,cla

y lo

am,2

%G

C17

28C

heng

and

Joh

nson

(19

98)

Whe

atA

mbi

ent C

O2,

No-

NM

ollis

ol,c

lay

loam

,2%

GC

4228

Che

ng a

nd J

ohns

on (

1998

)W

heat

Ele

vate

d C

O2,

+N

Mol

lisol

,cla

y lo

am,2

%G

C73

28C

heng

and

Joh

nson

(19

98)

Sunfl

ower

Am

bien

t CO

2M

ollis

ol,c

lay

loam

,2%

Mes

ocos

m55

53C

heng

et a

l. (2

000)

Sunfl

ower

Ele

vate

d C

O2

Mol

lisol

,cla

y lo

am,2

%M

esoc

osm

3153

Che

ng e

t al.

(200

0)Sp

ybea

nM

ollis

ol,c

lay

loam

,2%

GH

164

119

Che

ng e

t al.

(200

3)W

heat

Mol

lisol

,cla

y lo

am,2

%G

H96

119

Che

ng e

t al.

(200

3)So

ybea

nTa

llgra

ss p

rair

ie s

oil (

C4)

Mol

lisol

,cla

y lo

am,2

%G

H70

120

Fu a

nd C

heng

(20

02)

Sunfl

ower

Tallg

rass

pra

irie

soi

l (C

4)M

ollis

ol,c

lay

loam

,2%

GH

3912

0Fu

and

Che

ng (

2002

)So

rghu

mC

oast

al g

rass

land

soi

l (C

3)Sa

ndy

loam

,1.9

%G

H−9

120

Fu &

Che

ng (

2002

)A

mar

anth

usC

oast

al g

rass

land

soi

l (C

3)Sa

ndy

loam

,1.9

%G

H−5

120

Fu &

Che

ng (

2002

)W

heat

12/1

2 h

light

/dar

k cy

cle

Mol

lisol

,cla

y lo

am,2

%G

C10

038

Kuz

yako

v an

d C

heng

(20

01)

Whe

at12

/60

h lig

ht/d

ark

cycl

eM

ollis

ol,c

lay

loam

,2%

GC

−50

38K

uzya

kov

and

Che

ng (

2001

)

†Pla

nt g

row

th c

ondi

tions

:GH

= g

reen

hous

e; G

C =

gro

wth

cha

mbe

r.‡E

ach

valu

e is

cal

cula

ted

as:(

plan

ted—

unpl

ante

d)/u

npla

nted

×10

0.

07Cheng 5/20/05 13:34 Page 128



below the no-plant control after two long dark periods (60 h dark each, with a 12-h light period in between) were imposed. The study indicated that plant photo-synthesis controlled total rhizosphere respiration and exudation directly, therebyinfluencing the priming effect indirectly by changing the amount of easily de-composable organic substances in the rhizosphere.

Based on data from several experiments, timing of plant growth or plantphenological stages seems to be a very important factor influencing the magnitudeof the rhizosphere priming effect (Fig. 7–3) (Cheng et al., 2003; Fu and Cheng,2002). Rhizosphere priming effects seem low or even negative during the earlygrowth stages, increases to the highest point at approximately 60 d after plantingor around the flowering stage, and declines to lower levels afterwards. This is log-ical since the release of substrates in the rhizosphere is also controlled by the tim-ing of plant growth or plant phenological stages (Warembourg and Estelrich,2000).

MECHANISMS

As shown in the previous sections, plant roots may strongly influence SOMdecomposition. These root effects can be in the forms of decreasing mineral nu-trient availability to soil microorganisms due to plant uptake (Schimel et al., 1989;Schimel and Chapin, 1996), changing the physical and chemical environment inthe rhizosphere (i.e., water, pH, etc.) (Fuhr and Sauerbeck, 1968; Shields and Paul,1973; Jenkinson, 1977), increasing organic substrates (i.e., exudates, other rhi-

Fig. 7–3. Change of rhizosphere priming through time of wheat growth. The data points on the leftshorter than 40 d are from three separate experiments using growth chambers (Cheng, 1996; Chengand Johnson, 1998; Kuzyakov and Cheng, 2001), and the three data points on the right longer than40 d are from one greenhouse experiment (Cheng et al., 2003).

07Cheng 5/20/05 13:34 Page 129



zodeposition) for rhizosphere microorganisms, and enhancing microbial turnoverdue to faunal grazing (Ingham et al., 1985; Elliott et al., 1979; Clarholm, 1985a,1985b; Kuikman et al., 1990). However, direct mechanistic investigations of theseroot effects are still sparse.

There are controversies about the mechanisms of root effects on the inten-sity of SOM decomposition. Some hypotheses have been given in reconcilingthese controversies. Plant roots may have dual and counteracting effects on soilmicrobial activities and thereby SOM decomposition (Van Veen et al., 1989). Thefollowing mechanisms have been suggested to explain the effect of roots on SOMdecomposition:

• Drying effect hypothesis• Aggregate destruction hypothesis• Competition hypothesis• Preferential substrate utilization hypothesis• Microbial activation hypothesis• C uptake hypothesis

These mechanisms are presented and discussed below.

Drying Effect Hypothesis

Water uptake by plants always results in drier soil conditions in plantedtreatments than unplanted controls. Although frequent watering can be carried outto compensate the water difference between the two kinds of treatments, drying–rewetting cycles occur inevitably more in the planted treatment than the no-plantcontrol, especially in the topsoil (Sala et al., 1992). According to the results fromsome experiments investigating soil drying–rewetting cycles, the change in waterregime tends to induce an increased SOM mineralization (Van Schreven, 1967;Lundquist et al., 1999). This effect has been attributed to (i) enhanced solubilityof humic substances, (ii) increased microbial death during desiccation and os-motic shock caused by rewetting followed by an acceleration in decompositionand mineralization rates during microbial regrowth, and (iii) release of protectedorganic matter by disruption of macroaggregates during rewetting due to ‘slaking’(Magid et al., 1999). Since root hairs are responsible for water uptake as well asfor the exudation of readily available organic substances, these locations containhigh density of bacteria. Therefore the drying of soil particles on the root surfacecan also lead to dehydration of some microbial cells. Subsequent rewetting mayenhance the utilization of dead cells and an increased C and N mineralization (VanGestel et al., 1993). This mechanism may explain the increased CO2 efflux and Nmineralization in the presence of the rhizosphere. On the other hand, low soilmoisture may decrease microbial activities. So, under prolonged dry soil condi-tions (not a rapid change), the SOM decomposition must decrease. Therefore,under certain conditions the negative effect of water limitations on microbial ac-tivities and the positive effect of enhanced substrate availability (as mentionedabove) can be balanced out, resulting in little change in SOM decomposition whensoils are exposed to the drying-rewetting regime (Degens and Sparling, 1995).This possibility is clearly demonstrated by the results of a recent study (Magid et

07Cheng 5/20/05 13:34 Page 130



al., 1999), indicating that drying and rewetting a loamy sand soil did not signi-ficantly alter the rate of native SOM decomposition but reduced plant litter de-composition, as compared to the constant soil moisture regime. These balancingmechanisms in the drying–rewetting regime may lower the magnitude of the en-hanced SOM decomposition compared to the degree of rhizosphere priming ef-fect. However, the potential contribution of the drying-rewetting in the rhizo-sphere to the priming of SOM decomposition warrants further investigations.

Aggregate Destruction Hypothesis

Some studies have shown that a portion of labile SOM may be physicallyprotected from microbial utilization due to the formation of soil aggregates (Beareet al., 1994; Elliott, 1986; Tisdall and Oades, 1982). If the presence of live rootspromotes the destruction of aggregates rather than their formation, some portionof the physically protected labile SOM is further exposed to microbial attack,thereby resulting in positive priming effects of roots on SOM decomposition.Helal and Sauerbeck (1984, 1986) first put out this hypothesis to explain the pos-itive priming effect of live roots on SOM decomposition. According to this mech-anism, destroying aggregates increases the SOM mineralization. Indeed, rootspromote the reorganization of the soil structure: they can destroy some old aggre-gates as well as contribute to the formation of new ones. However, live roots tendto promote aggregate formation more than aggregate destruction (Haynes andBeare, 1997). Roots tend to occupy existing soil pores instead of creating newones. This feature of root growth is particularly prevalent in soils or horizons withfine texture such as clay or clayey loam. However, extreme drying–rewetting cy-cles, more likely to occur in the presence of live roots than no-root controls insome experiments, may result in aggregate destruction more than aggregate for-mation, thereby leading to priming effects on SOM decomposition. In most of thestudies cited in the above section, extreme drying–rewetting cycles were avoidedby frequent watering. Therefore, aggregate destruction may not have been themain cause of root priming effects on SOM decomposition. However, publishedstudies of root priming effects on SOM decomposition rarely include assessmentof root effects on soil aggregates simultaneously. Further studies on this issue areneeded before a general conclusion can be reached.

Competition Hypothesis

It is hypothesized that, when plants are grown in soils with low nutrient con-centrations, nutrient uptake by roots will intensify the competition for mineral nu-trients in the rhizosphere and decrease microbial growth and metabolism, therebydepressing SOM decomposition. A similar hypothesis was proposed by Cheng(1999) to explain how rhizosphere processes change when plants are grown underelevated CO2 conditions. This mechanism has been suggested indirectly bySchimel et al. (1989) and Ehrenfeld et al. (1997), but the hypothesis has neverbeen directly tested. The competition hypothesis has been used to explain the neg-ative effects of roots on SOM decomposition in several studies (e.g., Bottner et al.,1999). This hypothesis is based on the assumption that N is the most limiting nu-

07Cheng 5/20/05 13:34 1



trient for plants and microorganisms in terrestrial ecosystems (Vitousek andHowarth, 1991). Microbial growth in the rhizosphere is especially limited byavailable N as well as by some other nutrients (Schimel et al., 1989; Jackson et al.,1989; Liljeroth et al., 1990). Therefore, under conditions of limited soil mineralnutrient supply, plant roots and microorganisms compete strongly for available Nresources (Wang and Bakken 1997; Jackson et al., 1989). However, the winner inthis competition may depend on the time period considered. In the short-term(hours to days), soil microorganisms can capture ammonia (NH4

+) or nitrate(NO3

–) two to five times faster than the plant roots (Jackson et al., 1989), becausemicroorganisms have high substrate affinities, rapid growth rates, and high sur-face/volume ratios (Rosswall, 1982; Jackson et al., 1989). The short-term immo-bilization of N added in the form of mineral fertilizer is one of the frequently ob-served results of this competition (e. g., Bremer and Kuikman, 1997). However, inthe longer-term (weeks to months), plant roots may take up more N from the soileven in the presence of microbial competition. This competition mechanism canbe represented by a simple model (Fig. 7–4). Because soil microorganisms nor-mally grow and die (or turnover) at a much faster rate than do most plant roots,mineral nutrients taken up by microorganisms in the short-term are returned to themineral nutrient pool as microorganisms die and decompose, whereas nutrientstaken up by roots are mostly transported to other parts of the plant and little returnsto the soil mineral pool within a growing season. Root exudation of N containingcompounds usually does not exceed 5% of total plant N uptake (Merbach et al.,1999), which is one to two orders of magnitude lower than the amount of N min-eralized from microbial biomass turnover. Root turnover can be another source ofplant N loss to the soil within a growing season and should be a small portion ofthe total plant N uptake. Because of this difference in turnover rates between plantmaterials and soil microbial biomass, the flow between the microbial pool and themineral N pool is bi-directional and the flow between the mineral N pool to rootsis unidirectional. In such a system (Fig. 7–4), plants win the competition in termsof net N uptake in the longer-term, and the amount of net N uptake by microor-ganisms can be either positive, zero, or negative depending on other sources of Nand C substrates in the system. If the mechanism presented in this simple modeloperates widely in natural ecosystems in the longer-term, plants always win thecompetition for net N uptake as long as soil microorganisms turnover at a fasterrate. Under N limiting conditions, the removal of available N from the soil pool byroots may reduce microbial growth due to N limitation, resulting in a decreasedrate of SOM decomposition. This mechanism has been suggested by several stud-

Fig. 7–4. The box and arrow diagram represents a mechanism of competition for mineral nitrogen (N)between plant roots and soil microbes.

07Cheng 5/20/05 13:34 Page 132



ies (Jackson et al., 1989; Schimel et al., 1989; Kaye and Hart, 1997; Hodge et al.,2000a) and supported by the results of Hodge et al. (2000b). A similar hypothesishas also been used to explain the competition between soil microorganisms andforest plant communities (Hobbie, 1992). It has been shown that phosphorus (P)limitations may also reduce the rate of SOM decomposition (Hobbie and Vi-tousek, 2000). This competition hypothesis postulates that root effects on SOMdecomposition depend on the level of soil mineral nutrition: (i) under nutrient lim-iting conditions, SOM decomposition decreases due to competition between rootsand soil microorganisms for mineral nutrients; and (ii) the effect of roots on SOMdecomposition increases when soil mineral nutrients are abundant (Fig. 7–5).

Preferential Substrate Utilization Hypothesis

This hypothesis states that, given abundant mineral nutrient supply, soil mi-crobes prefer labile root-derived C to SOM-derived C, resulting in a decreasedSOM decomposition in the rhizosphere and that, if mineral nutrients are in shortsupply, soil microbes prefer nutrient-rich SOM to root-derived C, resulting in in-creased SOM decomposition in the rhizosphere (Fig. 7–6). A similar hypothesishas been given in a review (Cheng, 1999). This hypothesis focuses on the role ofsoil mineral nutrition and assumes that all root-derived materials have a muchwider C/N ratio than the SOM. The initial supporting data for this hypothesiscomes from some studies using the continuous 14C-labelling approach (Merckx etal., 1987; Van Veen et al, 1989; Liljeroth et al., 1994). The decomposition rate of14C-labelled materials from the current season rhizodeposition appeared to behigher under the treatment receiving N fertilization (or abundant mineral nutrientsupply) than under the unfertilized treatment (or mineral nutrient limited), and thedecomposition rate of original SOM decreased under the fertilized treatment.These results seem to support the idea that the microbial preference for substrate

Fig. 7–5. A graphical presentation of the competition hypothesis in relation to mineral nutrient levels.

07Cheng 5/20/05 13:34 Page 133



Fig. 7–6. A graphical presentation of the preferential substrate utilization hypothesis in relation to thelevel of soil mineral nutrients.

utilization switches depending on the level of soil mineral nutrients. This prefer-ential substrate utilization hypothesis seems contradictory to the competition hy-pothesis at the first glance but may be explained if both are shown on a gradient ofmineral nutrition (Fig. 7–7). Since the competition hypothesis is primarily basedon results from experiments in nutrient poor soils (pine forests in Ehrenfeld et al.[1997] and dry grasslands in Schimel et al. [1989]), it belongs to the lower part ofthe mineral nutritional gradient (or the very left part of the X-axis). The substrate-preference hypothesis can be placed to the right part of the X-axis because the ev-idence supporting the hypothesis mainly came from experiments using very fer-

Fig. 7–7. A reconciliation of the preferential substrate utilization hypothesis and the competition hy-pothesis.

07Cheng 5/20/05 13:34 Page 134



tile soils from agricultural fields. In other words, the mechanism of competitiondominates under the condition of severe mineral nutrient limitation, the substrate-preference mechanism dominates when soil mineral nutrients are not so limiting,and the two mechanisms may balance out in the intermediate level of soil mineralnutrients. Further work is needed to verify these possibilities.

Microbial Activation Hypothesis

This hypothesis has previously been described by Kuzyakov et al. (2000).Substances released by roots are readily available for rhizosphere microorgan-isms. These substrates normally stimulate microbial growth in the rhizosphere.This increased microbial growth may further lead to an increased co-metabolic de-composition of SOM, thereby resulting in an increased rate of SOM decomposi-tion. Microorganisms are the key players for producing this rhizosphere primingeffect, because no real priming effects have been observed under sterile conditions(Jansson, 1958). Also, the close relationship between microbial growth and the in-creased mineralization rate indicates a real connection between the dynamics ofmicrobial activities and real priming effects (Dalenberg and Jager, 1989). There isalso a significant correlation between the amount of N released from SOM and thelevel of exocellular enzyme activities, especially the total and soluble protease, asshown in an experiment studying the effect of glucose addition to the soil (Asmaret al., 1994). In experiments using soils of high N levels, Schmitt et al. (1991) re-ported that a pulse input of available substrates such as glucose enhances dehy-drogenase activity and increases the number of ammonifying and protolytic bac-teria but decreases the concentration of total organic C in the soil solution. Anincreased microbial substrate availability induces enzyme production or increasesenzyme activity, further leading to a co-metabolic decomposition of SOM.

Some studies have suggested that the response of total microbial metabo-lism determines the effect of roots on SOM decomposition (Fig. 7–8) (Cheng andColeman, 1990). Increased microbial biomass is reported when a stimulatory ef-fect of roots on SOM decomposition occurs (Helal and Sauerbeck, 1986; Sallihand Bottner, 1988; Cheng and Coleman, 1990). Decreased microbial biomass isindicated when a negative effect of roots on SOM decomposition is found (Reidand Goss, 1982; 1983; Sparling et al., 1982; Sallih and Bottner, 1988).

The input of labile root-derived C in the rhizosphere initially may decreaseSOM decomposition due to the increase of microbial growth and immobilization.But later it may stimulate SOM decomposition and nutrient release due to theturnover of this newly grown microbial biomass (Fig. 7–9). The quality of theroot-derived substrates is an important determinant of the timing and the magni-tude of the priming effect. This hypothesis emphasizes the temporal microbial dy-namics and the quality of the root-derived substrates. This hypothesis can be usedto potentially explain all the results mentioned above if the information on micro-bial dynamics and the quality of root exudates is available and correct. Unfortu-nately, such information is difficult to obtain and rarely available. Studies using14C-labelled substrates have also shown that the priming effect of added labilesubstrates on SOM decomposition is mainly due to the stimulation of microbialgrowth and subsequent microbial turnover (death) (Dalenberg and Jager, 1989;

07Cheng 5/20/05 13:34 Page 135



Fig. 7–8. Rhizosphere priming affects soil organic matter (SOM) decomposition by changing soil mi-crobial metabolism. If soil microbial biomass and metabolism are increased by the presence of therhizosphere, SOM decomposition is stimulated.

Fig. 7–9. A graphical presentation of the microbial activition hypothesis and the time course of thepriming effect.

Nicolardot et al., 1994). Microbial biomass turns over faster in the rhizospherethan in the bulk soil, often resulting from intensified predation by rhizosphericfauna (Elliott et al., 1988; Clarholm, 1985a; Ingham et al., 1985; Griffiths, 1994).Interactions between soil microorganisms, soil fauna, and roots are regarded asone of the keys for understanding SOM decomposition in the rhizosphere (Alpheiet al., 1996). Plant roots increase the microbial activity in the rhizosphere through

07Cheng 5/20/05 13:34 Page 136



rhizodeposition and thus can enhance N mineralization via activated foodweb in-teractions in the rhizosphere (Elliott et al., 1979; Clarholm, 1985b; Haider et al.,1987; Kuikman et al., 1991; Zagal, 1994; Zwart et al., 1994; Kuzyakov et al.,2001).

Carbon Uptake Hypothesis

If live roots take up SOM in a significant quantity, the SOM absorbed by thelive roots will be removed temporarily from microbial decomposition, resulting ina lower rate of SOM decomposition compared to SOM decomposition without thepresence of live roots. This mechanism has been suggested in some studies as apossible cause of the negative rhizosphere effect on SOM decomposition (Spar-ling et al., 1982; Reid and Goss, 1982; 1983). However, later studies indicate thatthe amount of soil C absorbed by roots is often <1% of the total decomposed SOM(as often measured by CO2 efflux) and is too small to be a significant factor influ-encing rhizosphere effects (Sallih and Bottner, 1988; Cheng and Coleman, 1990;Hodge et al., 2000a). Results from experiments using solution cultures under ster-ile conditions suggest that plant roots tend to absorb or re-absorb some quantity ofsoluble root exudates (Jones and Darrah, 1992, 1993, 1996) and limited quantityof dissolved organic N (e.g., amino acids) (Chapin et al., 1993; Kielland, 1994).However, the quantitative significance of such absorption under realistic condi-tions remains debatable. The rapid turnover of soluble organic substrates by soilmicroorganisms and the poor competitive ability of plant roots may constrain thequantity of root absorption of soluble organic materials to a relatively low level(Jones, 1999; Owen and Jones, 2001; Jones and Kielland, 2002).

Mechanism Interactions-A hypothesis

Based on the evidence obtained so far, the most important mechanisms con-trolling the rhizosphere effect on SOM mineralization are: preferential substrateutilization, competition for mineral nutrients, and microbial activation. In realitythese mechanisms may operate individually or in combination, and dominate de-pending on the availability of soil C and N at different spatial and temporal scales.Four scenarios can be given depending on which nutrient (C or N) is limiting themicrobial growth (Table 7–4). Microbial growth in non-rhizosphere soils is oftenlimited by the Cavailable (e.g., Wardle, 1992; Grayston et al., 1996), but, under cer-tain conditions, Cavailable in the rhizosphere may surpass the threshold of limitation

Table 7–4. Hypothesized dominating mechanism under different combination of carbon (C) andnitrogen (N) availability to microbial growth.

Soil conditions N limited N not limited

Cavailable limited Competition dominates Substrate preference dominates(negative rhizosphere priming) (negative rhizosphere priming)

Cavailable not limited Microbial activation or Microbial activation dominatesNutrient competition dominates (+ or � rhizosphere priming)(+ or � rhizosphere priming)

07Cheng 5/20/05 13:34 Page 137



to microbial growth (Cheng et al., 1996) and then Nmin becomes a limiting factor(Merckx et al., 1987; Van Veen et al., 1989). If soil microbial growth is limited byboth Cavailable and Nmin, the competition between plants and microorganisms forNmin dominates, leading to a decreased microbial activity, and, as a result, SOMdecomposition decreases. If only Cavailable limits microbial growth, microbes maypreferentially use rhizodeposits instead of SOM decomposition, then the mecha-nism of the preferential substrate utilization dominates. In the case that microbialgrowth is limited by Nmin but not limited by Cavailable, the rhizodeposition of easilyavailable organic compounds may either activate total microbial activity or pro-voke competition for mineral nutrients between roots and microbes, or both. Whenboth Cavailable and Nmin are not limiting microbial growth, the fast microbial growthinvokes an enhanced faunal predation in the rhizosphere, resulting in an increasedSOM decomposition and the domination by the microbial activation mechanism.

SUMMARY AND FUTURE RESEARCH

Based on results from experiments under laboratory or greenhouse condi-tions, root effects on the rate of SOM decomposition can range from negative 70%to as high as 330% above the unplanted control. It is clear that the rhizosphere sig-nificantly controls SOM decomposition. In all real terrestrial ecosystems, SOMdecomposition always occurs with plant roots, thereby inevitably entangles withboth the soil component and the plant component. This entanglement seriouslychallenges the reliability of existing assessments on the rates of SOM decomposi-tion that often come from measurements using root exclusions or soil incubationswithout the presence of plant roots (e. g., Bolker et al., 1998; Dalias et al., 2001;Parton et al., 1987).

Virtually all published data cited in this chapter come from experimentsusing herbaceous plant species. How tree roots may affect SOM decompositionremains to be investigated. Since forests constitute a major portion of the global Ccycle, understanding tree root effects on SOM decomposition may bear more sig-nificance in terms of quantifying the potential of C sequestration into forestecosystems at the global scale. Therefore, the effect of tree roots on SOM decom-position warrants future research.

Based on scattered studies, both biotic and abiotic factors, for example, N,CO2, soil moisture, light, soil types, plant species, and plant phenological stages,are found to significantly control or influence the level of root effects on SOM de-composition. Because rhizosphere processes are intimately connected with boththe plant system and the soil system (Högberg et al., 2001; Kuzyakov and Cheng,2001), any environmental conditions that affect either the plant functions or thesoil functions or both, inevitably modulate root effects on SOM decomposition.Comprehensive future studies are clearly needed to integrate all these importantfactors into a general model of understanding on this issue.

Studies of root effects on SOM decomposition have mostly been con-strained by the availability of research methods. What emerges from the discus-sion on methods is the need for the development of innovative methods that allowrealistic investigation of root effects on SOM decomposition in situ. Before such

07Cheng 5/20/05 13:34 Page 138



new methods become available, our understanding of root effects on SOM de-composition will have to rely on results from either greenhouse or laboratory stud-ies. Natural 13C tracers have been used in field-based studies for separating rhi-zosphere respiration from soil respiration (e.g., Rochette and Flanagan, 1997;Andrews et al., 1999). It is logical that these natural tracers can also be used forinvestigating root effects on SOM decomposition if soil moistures can be ade-quately controlled in both the rooted plots and the no-plant control plots.

The six mechanisms identified as potentially responsible for causing root ef-fects on SOM decomposition are preliminary in nature, and need to be rigorouslytested. For example, relevant to the drying effect hypothesis, the issue of differen-tial drying or wetting between the no-plant control and the planted treatments re-mains to be carefully investigated. Does frequent (e.g., daily) watering adequatelyeliminate the drying/wetting effect on SOM decomposition with or without roots?This potential drying/wetting effect may also disturb soil aggregate structures,thereby further influence SOM decomposition. The soil structure aspect of rooteffects on decomposition definitely needs more research before a general under-standing can be achieved. Microbial dynamics are either directly or indirectly in-volved in the six mechanistic hypotheses discussed above. Therefore, measure-ments of microbial community structure, growth and turnover, substrate use, andthe level of root-microbe associations are essential for reliable testing of these hy-potheses in the future.

Evidence from studies cited in this review supports a general belief that theroot effect on SOM decomposition can be large in magnitude and significant inmediating plant-soil interactions. However, very little is known about the role ofthe rhizosphere effect in shaping plant adaptation to various soil environments inthe long-term. If the rhizosphere effect are closely connected to plant photosyn-thesis and rhizodeposition (Högberg et al., 2001; Kuzyakov and Cheng, 2001), itis conceivable that the rhizosphere effect should be largely beneficial to plants, andthereby enhancing their fitness. Among possible benefits, enhanced nutrient ac-quisition is often suggested (Clarholm, 1985a; Ingham et al., 1985). Other bene-fits may include suppression of root pathogens by supporting healthy microbialcommunities (Hu et al., 1997), conditioning of soil paths for root growth, and im-proving soil structures and chemical environment such as pH adjustment(Schaller, 1987; Gahoonia and Nielsen, 1992). If all these benefits are true, the rhi-zosphere effect on SOM decomposition should be a result of evolutionaryprocesses operating between plants and soil organisms in the overall rhizospherecontinuum from incidental to highly symbiotic. Different rhizosphere mecha-nisms should be selected under different plant and soil environments. This argu-ment seems to be supported by the fact that different plant-soil couplings producedifferent rhizosphere effects on SOM decomposition (Fu and Cheng, 2002; Chenget al., 2003). Future research is needed to fully illuminate the evolutionary aspectsof the rhizosphere effect.

ACKNOWLEDGMENTS

We thank Ms. Cheryl Van De Veer and Dr. E. G. Gregorich for editorial as-sistance, and two anonymous referees for their constructive suggestions and qual-

07Cheng 5/20/05 13:34 Page 139



ity reviews. This research was supported by grants from the U.S. Department ofAgriculture (agreement no. 98-35107-7013), the National Science Foundation(DEB-9973995), and the German Research Foundation (DFG).

REFERENCES

Alphei, J., M. Bonkowski, and S. Scheu. 1996. Protozoa, Nematoda and Lumbricidae in the rhizo-sphere of Hordelymus europaeus (Poaceae): Faunal interactions, response of micro-organismsand effects on plant growth. Oecologia 106:111–126.

Andrews, J.A., K.G. Harrison, R. Matamala, and W.H. Schlesinger. 1999. Separation of root respira-tion from soil respiration using carbon-13 labeling during free-air carbon dioxide enrichment(FACE). Soil Sci. Soc. Am. J. 63:1429–1435.

Asmar, F., F. Eiland, and N.E. Nielsen. 1994. Effect of extracellular-enzyme activities on solubiliza-tion rate of soil organic nitrogen. Biol. Fertil. Soils 17:32–38.

Beare, M.H., M.L. Cabrera, P.F. Hendrix, and D.C. Coleman. 1994. Aggregate-protected and unpro-tected organic matter pools in conventional-tillage and no-tillage soils. Soil Sci. Soc. Am. J.58:787–795.

Bolker, BM., S.W. Pacala, and W.J. Parton, Jr. 1998. Linear analysis of soil decomposition: Insightsfrom the CENTURY model. Ecol. Appl. 8:425–439.

Bottner, P, M. Pansu, and Z. Sallih. 1999. Modelling the effect of active roots on soil organic matterturnover. Plant Soil 216:15–25

Bremer, E, and P. Kuikman. 1997. Influence of competition for nitrogen in soil on net mineralizationof nitrogen. Plant Soil 190:119–126

Chapin, F.S., L. Moilanen, and K. Kielland. 1993. Preferential use of organic nitrogen for growth bya nonmycorrhgizal arcitic sedge. Nature (London) 361:150–153.

Cheng, W. 1996. Measurement of rhizosphere respiration and organic matter decomposition using nat-ural 13C. Plant Soil 183:263–268.

Cheng, W. 1999. Rhizosphere feedbacks in elevated CO2. Tree Physiol. 19:313–320.Cheng, W., and D.C. Coleman. 1990. Effect of living roots on soil organic matter decomposition. Soil

Biol. Biochem. 22:781–787.Cheng, W., D.C. Coleman, and J.E. Box, Jr. 1990. Root dynamics, production and distribution in

agroecosystems on the Georgia Piedmont using minirhizotrons. J. Appl. Ecol. 27:592–604.Cheng, W., and D.W. Johnson. 1998. Effect of elevated CO2 on rhizosphere processes and soil organic

matter decomposition. Plant Soil 202:167–174.Cheng, W., D.W. Johnson, and S. Fu. 2003. Rhizosphere priming effects on decomposition: Controls

of fertilization, plant species, and phenology. Soil Sci. Soc. Am. J. 67:1418–1427.Cheng, W., D.A. Sims, Y. Luo, D.W. Johnson, J.T. Ball and J.S. Coleman. 2000. Carbon budgeting in

plant-soil mesocosms under elevated CO2: Locally missing carbon? Global Change Biol. 6:99–110.

Cheng, W., Q. Zhang, D.C. Coleman, C.R. Carroll and C.A. Hoffman. 1996. Is available carbon lim-iting microbial respiration in the rhizosphere? Soil Biol. Biochem. 28:1283–1288.

Clarholm, M. 1985a. Possible role for roots, bacteria, protozoa and fungi in supplying nitrogen toplants. p. 355–365. In A.H. Fitter (ed.) Ecological interactions in soil. Blackwell Science,Oxford.

Clarholm, M. 1985b. Interactions of bacteria, protozoa and plants leading to mineralization of soil ni-trogen. Soil Biol. Biochem. 17:181–187.

Dalenberg, J.W., and G. Jager. 1989. Priming effect of some organic additions to 14C-labelled soil. SoilBiol. Biochem. 21:443–448.

Dalias, P., J.M. Anderson, P. Bottner, and M.M. Coûteaux. 2001. Temperature responses of carbonmineralization in conifer forest soils from different regional climates incubated under standardlaboratory conditions. Global Change Biol. 7:181–192.

Degens, B., and G. Sparling. 1995. Repeated wet-dry cycles do not accelerate the mineralization oforganic C involved in the macro-aggegation of a sandy loam soil. Plant Soil 175:197–203.

Ehrenfeld, J.G., W.F.J. Parsons, X. Han, R.W. Parmelee, and W. Zhu. 1997. Live and dead roots in for-est soil horizons: Contrasting effects on nitrogen dynamics. Ecology 78:348–362.

Eissenstat, D.M., and R.D. Yanai. 1997. The ecology of root lifespan. Adv. Ecol. Res. 27: 1–60.Elliott, E.T. 1986. Aggregate structure and carbon, nitrogen, and phosphorus in native and cultivated

soils.. Soil Sci. Soc. Am. J. 50:627–633.

07Cheng 5/20/05 13:34 Page 140



Elliott, E.T., D.C. Coleman and C.V. Cole. 1979. The influence of amoebae on the uptake of nitrogenby plants in gnotobiotic soil. p. 221–229. In J.L. Harley and R.S. Russel (ed.) The soil-root in-terface. Academic Press, London.

Elliott, E.T., H.W. Hunt, and D.E. Walter. 1988. Detrital food-web interactions in North Americangrassland ecosystems. Agric. Ecosys. Environ. 24:41–56.

Epstein, H.E., I.C. Burke, and A.R. Mosier. 1998. Plant effects on spatial and temporal patterns of ni-trogen cycling in shortgrass steppe. Ecosystems 1:374–385.

Fu, S., and W. Cheng. 2002. Rhizosphere priming effects on the decomposition of soil organic matterin C4 and C3 grassland soils. Plant Soil 238:289–294

Fuhr, F., and D.R. Sauerbeck. 1968. Decomposition of wheat straw in the field as influenced by crop-ping and rotation. p. 241–250. In Isotopes and radiation in soil organic matter studies, IAEA,Vienna.

Gahoonia, T.S., and N.E. Nielsen. 1992. Control of pH at the soil-root interface. Plant Soil 140:49–54.Grayston, S.J., D. Vaughan, and D. Jones. 1996. Rhizosphere carbon flow in trees, in comparison with

annual plants: The importance of root exudation and its impact on microbial activity and nutri-ent availability. Appl. Soil Ecol. 5:29–56.

Griffiths, B.S. 1994. Soil nutrient flow. p. 65–92. In J.F. Darbyshire (ed.) Soil Protozoa. CAB Int.,Wallingford, Oxon, UK.

Haider, K., A. Mosier, and O. Heinemeyer. 1987. The effect of growing plants on denitrification athigh nitrate concentrations. Soil Sci. Soc. Am. J. 51:97–102.

Haynes, R.J. and M.H. Beare. 1997. Influence of six crop species on aggregate stability and some la-bile organic matter fractions. Soil Biol. Biochem. 29:1647–1653.

Helal, H.M., and D.R. Sauerbeck. 1983. Method to study turnover processes in soil layers of differentproximity to roots. Soil Biol. Biochem. 15:223–225.

Helal, H.M., and D.R. Sauerbeck. 1984. Influence of plant root on C and P metabolism in soil. PlantSoil 76:175–182.

Helal, H.M., and D.R. Sauerbeck. 1986. Effect of plant roots on carbon metabolism of soil microbialbiomass. Z. Pflanzenernaehr. Bodenkd. 149:181–188.

Helal, H.M., and D.R. Sauerbeck. 1987. Direct and indirect influence of plant roots on organic matterand phosphorus turnover in soil. INTECOL Bull. 15:49–58.

Hendrick, R.L., and K.S. Pregitzer. 1993. The dynamics of fine root length, biomass, and nitrogen con-tent in two northern hardwood systems. Can. J. For. Res. 23:2507–2520.

Hobbie, S.E. 1992 Effects of plant-species on nutrient cycling. Trends Ecol. Evol. 7:336–339.Hobbie, S.E., and P.M. Vitousek. 2000. Nutrient limitation of decomposition in Hawaiian forests.

Ecology 81:1867–1877Hodge, A., D. Robinson, and A. Fitter. 2000a. Are microorganisms more effective than plants at com-

peting for nitrogen? Trends Plant Sci. 5:304–308Hodge, A., J. Stewart, D. Robinson, B.S. Griffiths, and A.H. Fitter. 2000b. Competition between roots

and soil micro-organisms for nutrients from nitrogen-rich patches of varying complexity. J. Ecol.88:150–164.

Högberg, P., A. Nordgren, N. Buchmann, A.F.S. Taylor, A. Ekblad, M.N. Högberg, G. Nyberg, M. Ot-tosson-Lofvenius, and D.J. Read. 2001. Large-scale forest girdling shows that current photosyn-thesis drives soil respiration. Nature (London) 411:789–792.

Hu, S., A.H.C. van Bruggen, R.J. Wakeman, and N.J. Grünwald. 1997. Microbial suppression of invitro growth of Pythium ultimum and disease incidence in relation to soil C and N availability.Plant Soil 195:43–52.

Ingham, R.E., J.A. Trofymow, E.R. Ingham, and D.C. Coleman. 1985. Interactions of bacteria, fungiand their nematode grazers: Effects on nutrient cycling and plant growth. Ecol. Monogr. 55:119–140.

Jackson, L.E., J.P. Schimel, and M.K. Firestone. 1989. Short-term partitioning of ammonium and ni-trate between plants and microbes in an annual grassland. Soil Biol. Biochem. 21:409–415.

Jackson, R.B, H.A. Mooney, and E.D. Schulze. 1997. A global budget for fine root biomass, surfacearea, and nutrient contents. Proc. Natl. Acad. Sci. USA 94:7362–7366.

Janssens, I.A., H. Lankreijer, G. Matteucci, A.S. Kowalski, N. Buchmann, D. Epron, K. Pilegaard,W. Kutsch, B. Longdoz et al.. 2001. Productivity overshadows temperature in determining soiland ecosystem respiration across European forests. Global Change Biol. 7:269–278.

Jansson, S.L. 1958. Tracer studies on nitrogen transformations in soil with special attention to miner-alisation immobilisation relationship. Ann. R. Agric. Coll. Swed. 24:101–361.

Jenkinson, D.S. 1977. Studies on decomposition of plant material in soil.5: Effects of plant cover andsoil type on loss of carbon from C-14-labeled ryegrass decomposing under field conditions. J.Soil Sci. 28: 424–434.

07Cheng 5/20/05 13:34 Page 141



Jenny, H. 1941. Factors of soil formation: A system of quantitative pedology. McGraw-Hill, Mineola.NY.

Jones, D.L. 1999. Amino acid biodegradation and its potential effects on organic nitrogen capture byplants. Soil Biol. Biochem. 31:613–622.

Jones, D.L., and P.R. Darrah. 1992. Re-sorption of organic components by roots of Zea mays L. andits consequences in the rhizosphere. I. Resorption of 14C labelled glucose, mannose and citricacid. Plant Soil 143:259–266.

Jones, D.L., and P.R. Darrah. 1993. Re-sorption of organic components by roots of Zea mays L. andits consequences in the rhizosphere. II. Experimental and model evidence for simultaneous exu-dation and re-sorption of soluble C compounds. Plant Soil 153:47–59.

Jones, D.L., and P.R. Darrah. 1996. Re-sorption of organic components by roots of Zea mays L. andits consequences in the rhizosphere. III. Characteristics of sugar influx and efllux. Plant Soil 178:153–160.

Jones, D.L. and K. Kielland. 2002. Soil amino acid turnover dominates the nitrogen flux in per-mafrost-dominated taiga forest soils. Soil Biol. Biochem. 34:209–219.

Kaye, J.P., and S.C. Hart. 1997. Competition for nitrogen between plants and soil microorganisms.Trends Ecol. Evol. 12:139–143

Kielland, K. 1994. Amino acid absorption by arctic plants: Implications for plant nutrition and nitro-gen cycling. Ecology 75:2372–2383.

Kuikman, P.J., A.G. Jansen, and J.A. Van Veen. 1991. 15N-nitrogen mineralization from bacteria byprotozoan grazing at different soil moisture regimes. Soil Biol. Biochem. 23:193–200.

Kuikman, P.J., A.G. Jansen, J.A. Van Veen, and A.J.B. Zehnder. 1990. Protozoan predation and theturnover of soil organic carbon and nitrogen in the presence of plants. Biol. Fert. Soils 10:22–28.

Kuzyakov, Y., and W. Cheng. 2001. Photosynthesis controls of rhizosphere respiration and organicmatter decomposition. Soil Biol. Biochem. 33:1915–1925.

Kuzyakov, Y, and G. Domanski. 2000. Carbon input by plants into the soil: Review. J. Plant Nutr. SoilSci. 163:421–431.

Kuzyakov, Y., H. Ehrensberger, and K. Stahr. 2001. Carbon partitioning and below-ground transloca-tion by Lolium perenne. Soil Biol. Biochem. 33:61–74.

Kuzyakov, Y., J.K. Friedel, and K. Stahr. 2000. Review of mechanisms and quantification of primingeffects. Soil Biol. Biochem. 32:1485–1498.

Liljeroth, E., P. Kuikman, and J.A. Van Veen. 1994. Carbon translocation to the rhizosphere of maizeand wheat and influence on the turnover of native soil organic matter at different soil nitrogenlevels. Plant Soil 161:233–240.

Liljeroth, E., J.A. Van Veen, and H.J. Miller. 1990. Assimilate translocation to the rhizosphere of twowheat lines and subsequent utilization by rhizosphere microorganisms at two soil-nitrogen con-centrations. Soil Biol. Biochem. 22:1015–1021.

Lundquist, E.J., L.E. Jackson, and K.M. Scow. 1999. Wet-dry cycles affect dissolved organic carbonin two California agricultural soils. Soil Biol. Biochem. 31:1031–1038.

Lynch, J.M., and J. M. Whipps. 1990. Substrate flow in the rhizosphere. Plant Soil 129:1–10.Magid, J., C. Kjaergaard, A. Gorissen, and P. J. Kuikman. 1999. Drying and rewetting of a loamy sand

soil did not increase the turnover of native organic matter, but retarded the decomposition ofadded C-14-labelled plant material. Soil Biol. Biochem. 31:595–602.

Merbach, W., E. Mirus, G. Knof, R. Remus, S. Ruppel, R. Russow, A. Gransee, and J. Schulze. 1999.Release of carbon and nitrogen compounds by plant roots and their possible ecological impor-tance. J. Plant Nutr. Soil Sci. 162:373–383

Merckx, R.A., A. Dijkstra, A. den Hartog, and J.A. Van Veen. 1987. Production of root-derived mate-rials and associated microbial growth in soil at different nutrient levels. Biol. Fertil. Soils 5:126–132.

Nicolardot, B., G. Fauvet, and D. Cheneby. 1994. Carbon and nitrogen cycling through soil microbialbiomass at various temperatures. Soil Biol. Biochem. 26:253–261.

Owen, A.G. and D.L. Jones. 2001. Competition for amino acids between wheat roots and rhizospheremicroorganisms and the role of amino acids in plant N acquisition. Soil Biol. Biochem. 33:651–657.

Parton, W.J., D.S. Schimel, C.V. Cole, and D.S. Ojima. 1987. Analysis of factors controlling soil or-ganic matter level in Great Plains grasslands. Soil Sci. Soc. Am. J. 51:1173–1179.

Post, W.M., T-H Peng, W.R. Emmanuel, A.W. King, V.H. Dale, and D.L. DeAngelis. 1990. The globalcarbon cycle. Am. Sci. 78:310–326.

Pregitzer, K.S., D.R. Zak, P.S. Curtis, M.E. Kubiske, J.A. Teeri, and C.S. Vogel. 1995. AtmosphericCO2, soil nitrogen and turnover of fine roots. New Phytol. 129:579–585.

07Cheng 5/20/05 13:34 Page 142



Qian, J.H., J.W. Doran, and D.T. Walters. 1997. Maize plant contributions to root zone available car-bon and microbial transformation of nitrogen. Soil Biol. Biochem. 29:1451–1462.

Reid, J.B., and M.J. Goss. 1982. Suppression of decomposition of 14C-labelled plant roots in the pres-ence of living roots of maize and perennial ryegrass. J Soil Sci. 33:387–395.

Reid, J.B., and M.J. Goss. 1983. Growing crops and transformations of 14C-labelled soil organic mat-ter. Soil Biol. Biochem. 15:687–691.

Rochette, P., and L.B. Flanagan. 1997. Quantifying rhizosphere respiration in a corn crop under fieldconditions. Soil Sci. Soc. Am. J. 61:466–474.

Rosswall, T. 1982. Microbial regulation of the biogeochemical nitrogen cycle. Plant Soil 67:15–34Sala, O.E., W.K. Lauenroth, and W.J. Patron. 1992. Long-term soil water dynamics in the shortgrass

steppe. Ecology 73:1175–1181.Sallih, Z., and P. Bottner. 1988. Effect of wheat (Triticum aestivum) roots on mineralization rates of

soil organic matter. Biol. Fert. Soils 7:67–70.Schaller, G. 1987. pH changes in the rhizosphere in relation to the pH-buffering of soils. Plant Soil

97:439–444.Schimel, D.S. 1995. Terrestrial ecosystems and the carbon cycle. Global Change Biol. 1:77–91.Schimel, J.P., and F.S. Chapin, III. 1996. Tundra plant uptake of amino acid and NH4 nitrogen in situ:

Plants complete well for amino acid N. Ecology 77:2142–2147.Schimel, J.P., L.E. Jackson, and M K. Firestone. 1989. Spatial and temporal effects on plant microbial

competition for inorganic nitrogen in a California annual grassland. Soil Biol. Biochem 21:1059–1066.

Schmitt, L., K. Muller, and E. Ahrens. 1991. Chemical and microbiological changes in a Regosol aftermineral fertilization in long-term field experiments and short-term aerobic incubation trials. Z.Pflanzenernaehr. Bodenkd. 154:189–194.

Shields, J.A. and E.A. Paul. 1973. Decomposition of 14C-labelled plant material under field condi-tions. Can. J. Soil Sci. 53:297–306.

Smith, B.N., and S. Epstein. 1971. Two categories of 13C/12C ratios for higher plants. Plant Physiol.47:380–384.

Sparling, G.S., M.V. Cheshire, and C.M. Mundie. 1982. Effect of barley plants on the decompositionof 14C-labelled soil organic matter. J. Soil Sci. 33:89–100.

Tisdall, J.M., and J.M. Oades. 1982. Organic matter and water stable aggregates in soils. J. Soil Sci.33:141–161.

Van Gestel, M., R. Merckx, and K. Vlassak. 1993. Microbial biomass responses to soil drying andrewetting—The fate of fast growing and slow-growing microorganisms in soils from differentclimates. Soil Biol. Biochem. 25:109–123

Van Schreven, D.A. 1967. The effect of intermittent drying and wetting of a calcareous soil on carbonand nitrogen mineralization. Plant Soil 26:14–32.

Van Veen, J.A., E. Liljeroth, L.J.A. Lekkerkerk, and S.C. Van de Geijn. 1991. Carbon fluxes in plant-soil systems at elevated atmospheric CO2 levels. Ecol. Appl. 1:175–181.

Van Veen, J.A., R. Merckx, and S.C.Van De Geijn. 1989. Plant- and soil-related controls of the flowof carbon from roots through the soil microbial biomass. Plant Soil 115:179–188.

Vitousek, P.M., and R.W. Howarth. 1991. Nitrogen limitation on land and in the sea-How can it occur.Biogeochemistry 13:87–115.

Vogt, K.A., C.C. Grier, and D.J. Vogt. 1986. Production, turnover and nutrient dynamics of above- andbelowground detritus of world forests. Adv. Ecol. Res. 15:303–378.

Wang, J.G., and L.R. Bakken. 1997. Competition for nitrogen during decomposition of plant residuesin soil: Effect of spatial placement of N-rich and N-poor plant residues. Soil Biol. Biochem.29:153–162.

Wardle, D.A. 1992. A comparative assessment of factors which influence microbial biomass carbonand nitrogen levels in soil. Biol. Rev. 67:321–358.

Warembourg, F.R., and D.H. Estelrich. 2000. Towards a better understanding of carbon flow in the rhi-zosphere: a time-dependent approach using carbon-14. Biol. Fertil. Soils 30:528–534.

Zagal, E. 1994. Influence of light intensity on the distribution of carbon and consequent effects onmineralization of soil nitrogen in a barley (Hordeum vulgare L.)—Soil system. Plant Soil160:21–31.

Zak, D.R., and K.S. Pregitzer. 1998. Integration of ecophysiological and biogeochemical approachesto ecosystem dynamics. p. 372–403. In M.L. Pace and P.M. Groffman (ed.) Successes, limita-tions, and frontiers in ecosystem science. Springer-Verlag, New York.

Zwart, K.B., P.J. Kuikman, and J.A. Van Veen. 1994. Rhizosphere protozoa: Their significance in nu-trient dynamics. p. 93–122. In J.F. Darbyshire (ed.) Soil protozoa. CAB Int., Wallingford, Oxon,UK.

07Cheng 5/20/05 13:34 Page 143

Kitty’s TS • American Society of Agronomy • Wright & Zobel . . . • 18322507Cheng 5/20/05 13:34 Page 144

7 root effects on soil organic matter decomposition - …kuzyakov/cheng_asa_2005_roots-som.pdf ·...

Documents